How Emerging Optical Technologies will affect

the Future Internet

NSF Meeting, 5 Dec, 2005

Nick McKeownStanford University

nickm@stanford.eduhttp://www.stanford.edu/~nickm

Nick McKeown, 2005

2

Emerged (and deployed) Optical Technology

Well established Long-haul optical links Short-haul links between sub-systems

More recently deployed Photonic space switches Wavelength conversion

Q: Has optical technology affected the current Internet architecture?

Nick McKeown, 2005

3

Has optical technology affected the current Internet architecture?

Opinion #1: Not really. IP is oblivious to lower layers. IP will exploit any lower layer. Optics meant faster links: more of the same. Optics changed the numbers, but not the

architecture.

Opinion #2: Yes. Wildly. Imagine the Internet without optics. Abundant optical growth has transformed:

Topology, growth, scalability, usage, applications, and cost.

Nick McKeown, 2005

4

Emerging optical technology

1. Faster links

2. Lower cost and lower power Nanophotonics Integration of optics and electronics

• InP (e.g. single chip optical cross connects)• Silicon optics (e.g. SiGe modulators)

3. Optical packet switching Integrated optical processing, switching and

wavelength conversion Integrated optical packet buffers

Nick McKeown, 2005

5

Are faster optical links interesting?

Opinion #1: Who cares about links? We’ve moved to a period of abundance. Link bandwidth is no longer a constraint.

Opinion #2 Is abundance definitely the new order?

• Operators deliberately over-provision (e.g. fault recovery, and traffic growth; customers hate queues)

• Operators are losing money. • Is abundance sustainable?

Architecture is not oblivious to lower layers (e.g. wireless) Disruptive performance always disrupts the architecture

• Telephony: switching cost long-distance calls• Computer systems: Central timeshare mini desktop pda

My conclusion: Faster optical links will affect the Future Internet

Nick McKeown, 2005

6

Example of how optics can affect architecture Dynamic circuit switching in the backbone

Advantages of circuit switches Well-suited to optics Circuit switches are simple

“Start with a packet switch and throw most of it away”

Higher capacity per unit volume Higher capacity per watt Lower cost per Gb/s

Disadvantages They are unfashionable

Nick McKeown, 2005

7

DCS: Capacity on demand between border routers

Rule of thumb Predict the need for capacity by monitoring how quickly new

flows are created, rather than waiting for the buffer to fill

Rule of thumb Predict the need for capacity by monitoring how quickly new

flows are created, rather than waiting for the buffer to fill

Nick McKeown, 2005

8

My conclusion on dynamic circuit switching

Compelling to operator Cost, reliability, management, predictability

Scalable with optical circuit switching

Users can’t tell the difference

Prediction: The backbone will use some optical DCS in 10 years time

Nick McKeown, 2005

9

Emerging optical technology

1. Faster links

2. Lower cost, and lower power Nanophotonics Integration of optics and electronics

• InP (e.g. single chip optical cross connects)• Silicon optics (e.g. SiGe modulators)

3. Optical packet switching Integrated optical processing, switching and

wavelength conversion Integrated optical packet buffers

Nick McKeown, 2005

10

Integration of optics and electronics:Lower cost and lower power.

Effect on architecture of the last mile Very low-cost manageable optical switches on every

pole-top [Sandy Fraser].

Effect on architecture of routers Optical interconnects between chips, cards, shelves

and racks Higher bandwidth per unit volume Higher bandwidth per watt

General effect Integrated optics in 2005 are where integrated circuits

were in 1965 We can’t even imagine how optics will evolve

Nick McKeown, 2005

11

Emerging optical technology

1. Faster bit-pipes

2. Lower cost and lower power Nanophotonics Integration of optics and electronics

• InP (e.g. single chip optical cross connects)• Silicon optics (e.g. SiGe modulators)

3. Optical packet switching Integrated optical processing, switching and

wavelength conversion Integrated optical packet buffers

Nick McKeown, 2005

12

Optical Packet Switching

Conventional wisdom“A packet switch must...

1. Process headers, 2. Switch packet-by-packet, and3. Buffer packets during times of congestion.

Optics suck at all three.”

DARPA DOD-N Program revisiting assumptions(IRIS and LASOR projects)

1. Process headers: Carry headers slower; process electronically.2. Switch packets: Valiant Load Balancing (VLB) avoids packet-

by-packet switching [Sigcomm 03]

3. Buffer packets: 20-50 packets might be enough in the backbone [CCR 05]; will be feasible with integrated optics [Bowers 05]

Nick McKeown, 2005

13

Integrated optical buffers[Burmeister and Bowers, UCSB]

Think: 10-50 packets on a chip

Nick McKeown, 2005

14

Why we have big buffers today

Packet switching Long haul links are expensive Statistical multiplexing allows efficient sharing

of long haul links Packet switching requires buffers Packet loss is bad Use big buffers Luckily, big electronic buffers are cheap

Nick McKeown, 2005

15

Why bigger is not better

Network users don’t like buffers Network operators don’t like buffers Router architects don’t like buffers Optical buffers are very expensive

Electronics: Cheap buffers, expensive links Optics: Expensive buffers, cheap links

We don’t need big buffers

)(logO2

2 )2()1( Wn

CTCT

)(logO

22 )2()1( W

n

CTCT

1,000,000 10,000 20# packetsat 10Gb/s

Nick McKeown, 2005

16

Flexibility and Choice

Will it be optical DCS or optical packet switching? Technically, both seem feasible Perhaps we shouldn’t care Both are unfashionable Both should be on the table A new architecture should allow both… …but should presuppose neither

These are just examples. We should architect under the assumption that

both will be superseded